ExplainFix: Explainable Spatially Fixed Deep Networks

Abstract

Is there an initialization for deep networks that requires no learning? ExplainFix adopts two design principles: the "fixed filters" principle that all spatial filter weights of convolutional neural networks can be fixed at initialization and never learned, and the "nimbleness" principle that only few network parameters suffice. We contribute (a) visual model-based explanations, (b) speed and accuracy gains, and (c) novel tools for deep convolutional neural networks. ExplainFix gives key insights that spatially fixed networks should have a steered initialization, that spatial convolution layers tend to prioritize low frequencies, and that most network parameters are not necessary in spatially fixed models. ExplainFix models have up to 100x fewer spatial filter kernels than fully learned models and matching or improved accuracy. Our extensive empirical analysis confirms that ExplainFix guarantees nimbler models (train up to 17\% faster with channel pruning), matching or improved predictive performance (spanning 13 distinct baseline models, four architectures and two medical image datasets), improved robustness to larger learning rate, and robustness to varying model size. We are first to demonstrate that all spatial filters in state-of-the-art convolutional deep networks can be fixed at initialization, not learned.

arXiv commentRecently Published in Wiley WIREs Journal of Data Mining and Knowledge Discovery. This version has minor formatting differences and includes the supplementary appendix with the main document. Source code: https://github.com/adgaudio/ExplainFix/

References in corpus (19)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• Rethinking Atrous Convolution for Semantic Image Segmentation
• EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
• A Survey on Explainable Artificial Intelligence (XAI): Towards Medical XAI
• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• Multi-Scale Context Aggregation by Dilated Convolutions
• The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
• Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
• Explainable Artificial Intelligence: Understanding, Visualizing and Interpreting Deep Learning Models
• Understanding Deep Convolutional Networks
• Group Equivariant Convolutional Networks
• Gabor Convolutional Networks
• Deep Learning Scaling is Predictable, Empirically
• The Computational Limits of Deep Learning
• Scattering Networks for Hybrid Representation Learning
• Proving the Lottery Ticket Hypothesis: Pruning is All You Need
• Optimal Lottery Tickets via SubsetSum: Logarithmic Over-Parameterization is Sufficient
• Enhancement of Retinal Fundus Images via Pixel Color Amplification
• Influence-Directed Explanations for Deep Convolutional Networks